带有控制约束的不确定非线性系统的H_∞控制  

H_∞ Control of Uncertain Nonlinear Systems with Input Constraints

作　　者：崔小红[1] 王缔[1] CUI Xiao-hong WANG Di(Institute of Mathematical Sciences, Mudanjiang Normal College, Mudanjiang 157011, China)

机构地区：[1]牡丹江师范学院数学科学学院,黑龙江牡丹江157011

出　　处：《数学的实践与认识》2016年第22期215-223,共9页Mathematics in Practice and Theory

基　　金：牡丹江师范学院一般项目(YB201605)

摘　　要：研究了一类含有控制约束(饱和执行器)的不确定非线性系统的H_∞最优控制问题.首先,引入非二次型函数来解决控制受限的问题.然后借助二人零和微分对策理论研究H_∞最优控制问题,提出了模型自由的在线学习算法,仅仅通过在线获得的系统的状态信息而无需模型知识就能很好的解决H_∞控制问题.建立了执行网-评价网-扰动网三网络结构完成算法的在线实现.最后,利用李雅普诺夫(Lyapunov)稳定性理论给出了系统的稳定性分析,仿真算例验证了算法的有效性.H∞ optimal control for a class of uncertain nonlinear systems with input constraints（saturating actuators） is developed.Firstly,a nonquadratic function is introduced to handle with the problem of input constraints.Secondly,with the help of two-player zero-sum game theory,a model free online learning algorithm is designed to solve H∞ optimal control.The algorithm can solve the.H∞ optimal control effectively,which only needs the received measured state data without knowing the system dynamics.Critic-actor-disturbance neural network（NN） structure is built to establish the algorithm.Lastly,stability analysis is proved by Lyapunov theory,and the simulation example is given to demonstrate the effectiveness of the proposed algorithm.

关 键 词：神经网络 H∞控制 鲁棒控制 模型自由 

分 类 号：O232[理学—运筹学与控制论]